Detection and identification of islanding microgrid in active distribution networks using active frequency distribution and optimal utilization

• Introduces a novel tri-level framework that co-optimizes detection, resource scheduling, and network reconfiguration for islanding management. • Achieves superior performance with 0.931 detection sensitivity, sub-second detection, 42% less load shedding, and 12% lower operational costs. • Solves the complex MINLP problem via Benders decomposition with advanced linearization and an adaptive trust-region mechanism. • Establishes hard bidirectional coupling constraints that ensure detection feasibility given real-time operational and topological states. • Validates robustness and scalability through detailed 24-hour dynamic simulations on a modified IEEE 33-bus system. This paper presents a novel tri-level optimization framework that significantly reduces the impact of unintentional islanding in active distribution networks, achieving a 42% reduction in load shedding and a 12% decrease in daily operational costs compared to conventional passive detection methods. The approach maintains high detection reliability with sensitivity of 0.931, false alarm probability below 0.01, and sub-second detection times (0.68–0.92 s), while preserving power quality with voltage deviations within 0.02 p.u. and total harmonic distortion below 3.2%. The framework integrates active frequency perturbation, coordinated resource scheduling, and network reconfiguration into a single mathematically coupled model, moving beyond sequential or decoupled designs. The upper level optimizes perturbation parameters to maximize detection sensitivity while minimizing power quality degradation; the middle level schedules distributed energy resources to ensure stability and reserve adequacy; and the lower level manages network topology and load shedding to enable a seamless transition to islanded operation. Validated through detailed 24-hour dynamic simulations on a modified IEEE 33-bus system with multiple islanding events, the proposed co-optimization represents a foundational advance in holistic microgrid security management.